Temporal and spatial composition of the tumor microenvironment predicts response to immune checkpoint inhibition
Authors/Creators
-
Greenwald, Noah1
-
Nederlof, Iris2
- Sowers, Cameron1
-
Ding, Yi (Daisy)1
- Park, Seongyeol1
- Kong, Alex1
- Houlahan, Kathleen1
- Reddy Varra, Sri1
- de Graaf, Manon2
-
Liu, Candace1
-
Ranek, Jolene1
-
Voorwerk, Leonie2
- de Maaker, Michiel2
- Kagel, Adam1
-
McCaffrey, Erin
- Khan, Aziz1
-
Yeh, Christine
- Camacho Fullaway, Christine1
-
Khair, Zumana1
- Bai, Yunhao1
-
Piyadasa, Hadeesha3
-
Risom, Tyler4, 3
-
Delmastro, Alea
- Hartmann, Felix J5
- Mangiante, Lise1
-
Sotomayor-Vivas, Cristina6
- Schumacher, Ton N2
- Ma, Zicheng1
-
Bosse, Marc3
-
van de Vijver, Marc7, 8
-
tibshirani, robert
-
Horlings, Hugo
-
Curtis, Christina3
-
Kok, Marleen2
-
Angelo, Michael
-
1.
Stanford University School of Medicine
-
2.
The Netherlands Cancer Institute
-
3.
Stanford University
- 4. Genentech
-
5.
German Cancer Research Center
-
6.
Stanford Medicine
- 7. Amsterdam UMC Locatie VUmc
-
8.
Academic Medical Center
Description
Here we provide a spatiotemporal dataset of triple negative breast cance (TNBC), featuring matched primary tumors and longitudinal biopsies of metastatic lesions collected before and during ICI treatment (specifically anti-PD1) in the context of a prospective clinical trial. We generated multiplexed imaging data of pathology sections for each patient at different timepoints to conduct in-depth analysis of how spatial proteomic characteristics evolve through disease progression and immunotherapy in TNBC and their association with patient response.
All analysis scripts for this data can be found here.
Top Level Folders
analysis_files: This directory should initially contain a cell table (generated with Mesmer and annotated by Pixie). The scripts expect a column named "cell_meta_cluster" containing the cell clusters, as well "fov" with the specific image name. This folder will also contain the final data tables generated by the TNBC scripts.
output_files: This directory stores the per core and per timepoint data files for each feature. These are aggregated to form the final data tables stored in analysis_files.
intermediate_files: This directory contains subfolders storing any fov and cell level feature analysis done on the data. In addition, there is a subdirectory containing the metadata about each fov, each timepoint, and each patient.
Directory Tree
- TONIC_Cohort (base directory)
- analysis_files
- output_files
- intermediate_files
- metadata
- post_processing - contains specifications for the filtering of the data tables in output_files
- fiber_segmentation_processed_data - image level fiber analysis (code)
- tile_stats_512 - 512x512 tile analysis
- spatial_analysis
- neighborhood_mats - neighboring cell count/frequency at specified pixel radius and cell cluster level
- mixing_score - image level mixing score of various cell population combinations (code)
- cell_neighbor_analysis - data detailing cell diversity and linear distance between cell populations in an image (code)
- neighborhood_analysis - kmeans neighborhood analysis (code)
- ecm - generated in 4_ecm_preprocessing.py
- ecm_pixel_clustering - generated in ECM_Pixie_Cluster_Pixels.ipynb and ECM_pixel_clustering_stats.ipynb
Files
TONIC_Cohort.zip
Files
(5.9 GB)
| Name | Size | Download all |
|---|---|---|
|
md5:92168fd759061bfb8e8ea9f6f3c8c98a
|
5.9 GB | Preview Download |
|
md5:1b88abd4c90c94ffebd602728611f12b
|
285.8 kB | Preview Download |